Skip to main content

Filtered — This model refuses requests beyond its own limits outright. Measured, not assumed — models we have not measured carry no label at all.Browse every model at this filtering level
openai/gpt-image-2.5-sunburst/text-to-image

OpenAI's GPT Image 2.5 Sunburst text-to-image, the quality-first model of the GPT Image 2.5 pair. Choose five quality tiers from low to max, 1K, 2K or 4K output and 15 aspect ratios from 1:3 to 3:1, as PNG, JPEG or WebP.

From $0.0018/image20% below the publisher's list price
Run model

Workbench

Input

Text to image
315/5000
Up to 5,000 characters. Put every word that must appear in the image in double quotes and say where it goes; the rest is read as description. The first 100 input tokens of the prompt are included in the image price; longer prompts add a small charge per extra token, shown in the quote. Chinese, Japanese and Korean text counts about one token per character, other text about one token per four characters. at least 1 characters · at most 5000 characters
affects price
Fifteen shapes from 1:3 to 3:1, 1:1 by default. The shape changes the price: a wide or tall frame costs less than a square at the same tier.
affects price
1k (default), 2k or 4k. A square frame is 1024, 2048 or 2880 pixels across; a larger tier costs more per image. On the economy route a 1k image can come out somewhat larger.
Qualityaffects price
lowmediumhighxhighmax
low, medium, high (default), xhigh or max. Each step adds detail, time and cost; keep xhigh and max for finals.
Output format
PNGJPEGWebP
png (default), jpeg or webp. Does not change the price.
Output quality
Compression quality for jpeg and webp, 0 to 100: higher keeps more detail and makes larger files. Not used for png, and it does not change the price. no lower than 0 · no higher than 100
Official fallback
On by default. A request for a single high-quality 1k PNG in 1:1, 3:2, 2:3, 4:3, 3:4, 16:9 or 9:16 runs on the economy route first; if that route cannot complete it, for example because it refuses the prompt, the request runs again on the official route and is billed at the official price instead of the economy price. Every other request runs on the official route from the start. Turn it off to pay only the economy price: requests the economy route cannot serve are then rejected, and one it takes may still be refused or fail.
Safety checkerOff

SpicyAPI does not run a safety checker on this model. Your prompt and media go to the model as written, and the model itself decides what comes back.

Checking your account before enabling submission and pricing.

Output

Ready for a new run
ballroom-coupleExample result
Animate
Edit
Split into layers

Missing the model you need?

Tell us what to add and where you would use it. Requests go to the people who decide what gets onboarded next.

OPENAI IMAGE API

GPT Image 2.5 Sunburst API at a glance

OpenAI's precision-first GPT Image 2.5 Sunburst with five quality tiers up to max; high-quality 1K images start well below OpenAI's list price.

Economy route
1K
High-quality 1K images in seven common shapes start on a lower-priced route; Official fallback reruns anything it cannot finish on the official route.
Top quality tier
max
Five tiers: low, medium, high, xhigh and max.
Edit sources
16
Up to sixteen source images plus a mask in one edit.
Transparent output
Alpha
Cut-outs come back as PNG or WebP with a real alpha channel.

Results

What you can make with GPT Image 2.5 Sunburst

  • A man in a black tuxedo and a woman in a backless emerald satin gown dancing in an empty candlelit ballroom under crystal chandeliers

    Detail

    Fabric and light at the high tier

    A couple in a tuxedo and an emerald satin gown in an empty candlelit ballroom. At high quality the satin folds, the chandeliers and the reflections in the parquet hold up at full size.

    Text to image, high quality, 3:2.

  • The candlelit ballroom with its floor turned into a sheet of still water holding floating candles and reflecting the dancing couple

    Masked editing

    Repaint only the area you mark

    The same ballroom with a mask over the floor: the parquet became still water with floating candles, and the couple above the mask was left untouched.

    Edit with a mask over the floor.

  • The same woman and chair, her gown now crimson satin with the black lace sleeves kept

    Precision edits

    One garment changed, everything else kept

    The velvet gown turned into crimson satin with the lace sleeves preserved, while her face, the chair and the single key light did not move.

    Edit, one source, input fidelity high.

Content test

Is GPT Image 2.5 Sunburst uncensored? What we tested

We ran each level on our live setup and checked every clip frame by frame. A pass means the content was actually on screen, not just that the request went through.

Content levels tested on GPT Image 2.5 Sunburst, by mode
LevelImage editingText to image
MildThe lightest prompts in the test
RefusedTested Sep 29, 2026
RefusedTested Sep 29, 2026
ModerateWhere filtered models start to refuse
Not tested yet
Not tested yet
StrongThe middle of the scale
RefusedTested Sep 29, 2026
RefusedTested Sep 29, 2026
Very strongWhere most models start to hold back
Not tested yet
Not tested yet
MaximumThe top of the scale
Not tested yet
Not tested yet

Renders: on screen as written. Softened: generated, but milder than what was written. Refused: rejected. Not tested yet: we have not run that level on that mode.

How each mode behaves

All modes
OpenAI's safety system screens every prompt and result. In our tests every request from the Mild level up was refused, and in editing even a small wardrobe change toward that level was refused. Evening wear, fashion and portraits went through. Refused requests are not charged.

Use cases

What to build with the GPT Image 2.5 Sunburst API

  • Hero and product imagery

    Campaign key visuals and product shots viewed full screen, where the max tier's extra detail in fabric, glass and skin is worth its cost.

  • Precise retouching

    Wardrobe colour changes, relighting and masked fixes on approved photographs, where everything outside the change has to stay identical.

  • Cut-outs for layouts

    People and products on transparent backgrounds, ready to drop into a layout without a separate background-removal step.

Prompts

GPT Image 2.5 Sunburst prompt examples you can copy

Paste one into the playground as it is, then change one thing at a time.

  • Text to image

    Fabric and light at the high tier

    A man in a black tailored tuxedo and a woman in a backless emerald satin gown dance close in an empty candlelit ballroom, his hand at the small of her back, her head turned towards the camera over his shoulder. Crystal chandeliers, polished parquet reflecting the flames, shallow depth of field, fine fabric detail.

    Why it works. High quality at 3:2. The prompt spends its words on materials and light (satin, polished parquet, candle flames), which is where the higher tier shows.

    Try this prompt

  • Image editing

    Repaint only the area you mark

    In the masked lower part of the frame, turn the parquet floor into a shallow sheet of still water with small floating candles, reflecting the couple and the chandeliers. Leave the couple and the room above the floor untouched.

    Why it works. The mask is a PNG the same size as the source, transparent over the floor only. The prompt describes the new floor and says the couple and the room must stay untouched.

    This mode starts from your own image: add it in the playground before you run.

    Try this prompt

  • Image editing

    One garment changed, everything else kept

    Change the black velvet gown in Image 1 to a deep crimson satin gown that keeps the same long lace sleeves. Keep her face, hair, lipstick, pose, the carved chair, the backdrop and the single key light exactly as they are.

    Why it works. Only the gown colour and fabric are named as changes; the lace sleeves, face, chair and key light are listed as fixed.

    This mode starts from your own image: add it in the playground before you run.

    Try this prompt

Pricing

GPT Image 2.5 Sunburst API pricing

One image starts at $0.0018 at 1K and $0.0037 at 4K.

The playground shows the exact price for your settings before you run. Failed or timed-out tasks are usually not charged. Some models charge for content refusals as stated on their model page; check those billing terms before submitting.

See every price tier

Get started

How to use the GPT Image 2.5 Sunburst API in three steps

  1. Step 1

    Start at high, go to max for finals

    High is the default and costs a quarter of max at the same size. Use xhigh or max for hero images where fabric, hair and small type are looked at closely.

  2. Step 2

    Protect what must not change

    In edits, state the change and then everything that stays: face, pose, light. Add input fidelity high when the source is a person who has to stay recognisable.

  3. Step 3

    Ask for transparency by name

    Set background to transparent and output_format to png or webp, and say in the prompt what should be cut out. The result carries a real alpha channel.

Open the playgroundGet an API key

Parameter reference and code samples are in the API and Readme tabs at the top of this page.

Why SpicyAPI

Why run GPT Image 2.5 Sunburst on SpicyAPI

  • No filter from us

    We add no content filter of our own. What the model renders is what you get back.

  • Tested, not guessed

    Content labels come from real generations we ran and checked, with the test date on the page.

  • Pay per result, in dollars

    No subscription and no credits. Failed or timed-out tasks are usually not charged. Some models charge for content refusals as stated on their model page; check those billing terms before submitting.

  • One key for every model

    One API key and balance cover image, video, audio and chat models, so you can explore more models from the same account.

FAQ

GPT Image 2.5 Sunburst API FAQ

About GPT Image 2.5 Sunburst

GPT Image 2.5 Sunburst is OpenAI's precision model: five quality tiers up to max, transparent cut-outs, masked editing with up to sixteen sources and fifteen aspect ratios from 1K to 4K, with every setting priced below OpenAI's list price. OpenAI’s content policy applies, so it serves product, fashion and editorial work rather than work past OpenAI's policy line.

All GPT Image models3

Related models and next tasks